Abstract

Developing reading ability is an essential part of language acquisition. One of the core tasks is finding appropriate reading material for language learners at specific levels of language proficiency, which often proves challenging for both English instructors and readers. To automate the process of reading material selection and of reading ability assessment, a system that focuses on text readability analysis for second language (L2) learners is proposed.

The first part of the talk will present a dataset of CEFR -graded texts tailored for L2 learners of English and present a system aimed at automated text readability assessment for both native and L2 learners. We will address techniques used in the system (such as methods of generalisation, domain adaptation, and self-learning techniques) and discuss further challenges in this field. In the second part of the talk, we will present a prototype system aimed at automatically finding texts at appropriate levels of readability, and we will discuss the design of the automatically generated tasks to test learners’ reading proficiency.